Network Topologies

🧭

On this page — click any heading below to jump straight to it. ⬅️ Back to P-II.II. Computer Communications & Networks

---

OSI Layer: Physical Layer (L1)

Definition

Topology describes the physical layout (how cables/devices are actually arranged) or logical layout (how data actually flows, which can differ from the physical wiring) of nodes and links in a network. Topology choice drives cost, fault-tolerance, ease of installation, and scalability — it is one of the four axes used to classify networks alongside scale, transmission technology, and ownership.

Types of Network Topology

1. Bus Topology

• All nodes are tapped onto a single shared backbone cable (coaxial, historically), terminated at both ends by a terminator resistor to absorb signal and prevent reflection.
• Data broadcasts along the whole cable; every node "hears" every transmission but only the addressed node accepts it.
• Advantages: cheapest to install, minimal cabling, easy to extend for a small network.
• Disadvantages: a single break anywhere disables the entire segment; performance degrades sharply as nodes are added (shared bandwidth, collisions); difficult to troubleshoot a fault's exact location.
• Example: legacy 10BASE2/10BASE5 Ethernet.

2. Star Topology

• Every node connects by its own dedicated link to a central device (hub or switch); nodes never connect directly to each other.
• Advantages: a single cable failure only isolates one node, not the network; easy to install, extend, and troubleshoot (fault isolated to one link); centralised management.
• Disadvantages: the central hub/switch is a single point of failure for the whole network; more cabling than bus.
• Example: virtually all modern Ethernet LANs (wired to a switch).

3. Ring Topology

• Each node connects to exactly two neighbours, forming a closed loop; data travels around the ring hop by hop.
• Unidirectional ring: data flows in one direction only — simple but one broken link can down the whole ring.
• Bidirectional / dual ring: data can flow either way, and a second counter-rotating ring provides automatic failover if one link breaks (e.g., FDDI).
• Access is typically controlled by token passing — a node may transmit only while holding a special token frame, which eliminates collisions.
• Advantages: no collisions with token passing; performance is more predictable under heavy load than bus/Ethernet-style contention.
• Disadvantages: a single node/link failure can disrupt the whole ring (mitigated by dual-ring designs); adding/removing a node briefly disrupts the ring; overall delay grows with the number of nodes.
• Example: Token Ring, FDDI (Fibre Distributed Data Interface).

4. Mesh Topology

• Full mesh: every node has a dedicated link to every other node — maximum redundancy, requires n(n-1)/2 links for n nodes.
• Partial mesh: only critical/high-traffic nodes are fully interconnected; others connect through fewer links — a practical trade-off between mesh's reliability and its cost.
• Advantages: highest fault tolerance (loss of one link never isolates a node from the whole network), no traffic bottleneck/congestion at any single point, inherently private and secure per link, easy to identify a fault (isolated to one link).
• Disadvantages: extremely expensive and hard to install/reconfigure at scale — cabling and I/O ports grow quadratically (O(n²)) with the number of nodes.
• Example: core routers of the Internet backbone (partial mesh); critical military/banking links (full mesh).

5. Tree (Hierarchical) Topology

• A hybrid of bus and star: groups of star networks are connected to a bus backbone, forming a hierarchy with a root node and branching levels.
• Advantages: scalable — new branches added without disturbing the rest; fault isolated to one branch; supports point-to-point wiring per branch, standard for large organisations.
• Disadvantages: heavy reliance on the backbone/root — a break there can partition the whole tree; more complex to configure than a plain star.
• Example: a university's departmental LANs feeding into a central campus backbone.

6. Hybrid Topology

• Any deliberate combination of two or more basic topologies to exploit their respective strengths (e.g., star-bus, star-ring).
• Advantages: flexible and scalable, can be engineered to match specific reliability/cost needs of different network zones.
• Disadvantages: more complex design, installation, and management; troubleshooting requires understanding multiple topology rules at once.
• Example: a corporate campus with star-wired floors, each floor's switch uplinked to a ring or mesh core backbone.

Comparison Table

Numericals

Q1 — Links needed for full mesh

For n nodes, full mesh needs n(n-1)/2 links and each node needs n-1 I/O ports.

• For n = 7: links = 7×6÷2 = 21 links; each node needs 6 ports.

Q2 — Adding nodes: mesh vs ring vs bus

Existing network of 5 devices; 2 new devices are added.

• Mesh: before = 5×4÷2 = 10 links; after = 7×6÷2 = 21 links → 11 new links.
• Ring: before = 5 links (one per node); after = 7 links → 2 new links.
• Bus: only new drop cables to the shared backbone are needed → 2 new connections, no change to existing wiring.
• Conclusion: mesh scales as O(n²), ring and bus scale as O(n) — this is the standard exam argument for why full mesh is reserved for small, critical cores only.

CSS Exam Angle

• Draw a simple diagram for each topology — examiners award dedicated marks for diagrams, not just text.
• Structure the answer type-by-type with explicit advantages vs disadvantages, then close with the comparison table — this format alone often satisfies a full 8–10 mark question.
• Always mention the single point of failure concept where relevant (star's hub, tree's root, bus's backbone) — it is the recurring theme examiners probe.
• If marks allow, add the mesh-links numerical; it demonstrates quantitative understanding beyond memorised definitions.

Example / Pakistan Context

A university's departmental LANs (star-wired per lab) connected to a central campus backbone forms a tree topology. A bank's core-to-branch redundant links, engineered so no single fibre cut isolates a branch, resemble a partial mesh for high availability. Telecom exchange interconnects between major cities historically used ring topology (e.g., SDH/SONET rings) for automatic failover.

Likely Questions

1. Explain different network topologies with diagrams, highlighting their advantages and disadvantages.
2. Why is full mesh topology rarely used for large networks despite being the most fault-tolerant? Support with a numerical.
3. Differentiate bus, star, and ring topology in terms of fault tolerance and cost.
4. What is a hybrid topology? Give a real-world example.

📜 Past-Paper Questions & Structured Answers

Questions are paraphrased; answer focus is suggested guidance, not an official marking scheme.

2016 · Paper-II · Q2(a) · 6 marks

Question (paraphrased): A network grows from five to seven devices: calculate additional links for a full mesh and a ring.

Answer focus: Give each formula, old/new totals and the meaning of additional cables.

Full mesh: every device connects to every other device. An undirected n-node mesh requires n(n−1)/2 links.

• Initially: 5×4/2 = 10 links.
• Finally: 7×6/2 = 21 links.
• Additional links = 21−10 = 11. Check: each new node joins the five old nodes (10 links), plus one link between the new nodes.

Ring: a simple n-node ring contains n links. It grows from 5 to 7 links, so the net increase is 2 links.

For physical installation, break one existing link A—B and replace it with A—X—Y—B: three links are installed and one removed. Thus distinguish two additional links overall from three newly installed cable sections if the examiner interprets the wording physically.